Home

A RAID is a Redundant Array of Inexpensive Disks or sometimes Redundant Array of Individual Disks.

There are probably two main reasons behind a RAID, speed and security.  There are also many variations of RAID giving different levels of security and speed.

CnW Recovery currently handles RAID 0 which is a striped set of 2 disks. In this mode data is written in blocks, such as 128 sectors to each disk in turn. The benefit can be increase in speed.  The downside is there is no error recovery built in, so statistically a RAID 0  is twice as likely to fail as a normal drive.  A failure of a single drive, then often means that the complete RAID has failed.

RAID 1 is a complete mirrored drive.  ie two identical drives, and typically no speed benefit, but a single drive failure means no loss of data.

Higher RAID systems have a mixture of redundancy and speed benefits and the design goals are to get complete security from a single disk failure, but at the same time requiring fewer than double the number of drives, as in RAID 1.  Data is therefore split over several drives in a way that any failure will allow all data to be reconstructed.  This feature is only supported by the RAID option within CnW Recovery software

CnW Software handles RAID 0 Striped disks via the Image disk function.  An image file is created with the RAID option enabled.  Both disks need to be read and it is important that it is determined which is Disk 0 and which is Disk 1.  It is also important that the strip size is determined and set.  Both disks are then imaged to the same output drive and data is placed in the correct logical location. A small problem with this approach may be the necessity to have a temporary drive with the capacity to store an image of both disk drives.  Once the image has been created, it is treated as a simple disk image - the RAID element has in effect been removed

As with standard imaging, the image need not be done in a single go, thus areas of a bad disk may be skipped and data will still be placed in the correct logical location.  Thus a RAID with a errors in one area on one drive can largely be recovered.

Interleave size

To recreate a logical image of the interleave is critical.  Typical values are 0x80 or 0x100 (128 or 256) sectors.  Determining the value does require looking at sectors to work out which physical location they should be in and seeing it matches a possible interleave value.  The choice of disk is much easier, disk 0 will always store the start of the disk, so expect to see a boot sector at the start of disk 0 and not at the start of disk 1.

For an NTFS disk, the MFT often starts at 0x60003F, so for a RAID0 this will be 0x30003F.  It is then a matter of looking at the values of each MFT